Quality
Known Good Index and the capability indices, side by side
Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.
| Metric | Anthropic: Claude Opus 4.6 | OpenAI: gpt-oss-120b |
|---|---|---|
| Known Good Index | 87 | 62 |
| Coding Index | 92 | — |
| Reasoning Index | 77 | — |
| Openness Index | 5 | 60 |
Quality Breakdown
Every evaluation both models report a published score for
Bar = Anthropic: Claude Opus 4.6 · marker = OpenAI: gpt-oss-120b · scores ingested from Epoch AI (CC-BY 4.0)
Right-hand figures read Anthropic: Claude Opus 4.6 / OpenAI: gpt-oss-120b. Anthropic: Claude Opus 4.6 leads on 3 of 3, OpenAI: gpt-oss-120b on 0.
Arena Elo
Human preference rating from blind pairwise votes
Elo from blind pairwise votes, ingested from Arena. Higher is better.
| Metric | Anthropic: Claude Opus 4.6 | OpenAI: gpt-oss-120b |
|---|---|---|
| Text Arena | 1,498 | 1,365 |
| Text Arena (style controlled) | 1,497 | 1,352 |
Price & Cost
Published API pricing by token type
USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.
| Metric | Anthropic: Claude Opus 4.6 | OpenAI: gpt-oss-120b |
|---|---|---|
| Input $/1M | $5.00 | $0.04 |
| Output $/1M | $25.00 | $0.17 |
| Blended $/1M | $10.00 | $0.07 |
| Cache read $/1M | $0.500 | — |
Context Window
Maximum input accepted and maximum output emitted
Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.
| Metric | Anthropic: Claude Opus 4.6 | OpenAI: gpt-oss-120b |
|---|---|---|
| Context window | 1.00M | 131k |
| Max output tokens | 128k | 118k |
Speed & Latency
Mean throughput and time to first token across hosting providers, over successful measurements from the last 48 hours
Means across all tracked providers over successful measurements from the last 48 hours, derived from OpenRouter endpoint telemetry.
| Metric | Anthropic: Claude Opus 4.6 | OpenAI: gpt-oss-120b |
|---|---|---|
| Output tokens/s | 38 | 151 |
| Time to first token | 2.0s | 0.8s |
Providers
Which hosts serve each model, and at what price and speed
| Provider | Anthropic: Claude Opus 4.6 blended | Anthropic: Claude Opus 4.6 tok/s | OpenAI: gpt-oss-120b blended | OpenAI: gpt-oss-120b tok/s |
|---|---|---|---|---|
| Google · first party | $10.00 | 40 | $0.16 | 108 |
| Amazon Bedrock · first party | $10.00 | 39 | $0.26 | 290 |
| Claude Platform on AWS | $10.00 | 39 | — | — |
| Anthropic · first party | $10.00 | 36 | — | — |
| Azure · first party | $10.00 | 30 | — | — |
| Cerebras | — | — | $0.45 | 828 |
| Groq | — | — | $0.26 | 323 |
| SambaNova | — | — | $0.34 | 264 |
| Nebius | — | — | $0.26 | 259 |
| Crusoe | — | — | $0.10 | 190 |
| BaseTen | — | — | $0.20 | 176 |
| Mara | — | — | $0.30 | 172 |
| Novita | — | — | $0.10 | 138 |
| Parasail | — | — | $0.26 | 119 |
| Phala | — | — | $0.26 | 116 |
| DeepInfra | — | — | $0.07 | 114 |
| Together | — | — | $0.26 | 88 |
| AkashML | — | — | $0.07 | 61 |
| DigitalOcean | — | — | $0.15 | 44 |
| Mancer 2 | — | — | $0.10 | 43 |
| CoreWeave | — | — | $0.07 | 37 |
| SiliconFlow | — | — | $0.26 | 24 |
| DekaLLM | — | — | $0.07 | 22 |
Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.